# Re: st: Problem outregging GLLAMM results

 From Sophia Rabe-Hesketh To statalist@hsphsun2.harvard.edu Subject Re: st: Problem outregging GLLAMM results Date Mon, 19 Apr 2004 10:33:16 -0700

```Matt,

The problem with the number of observations
was due to gllamm storing the number of clusters
in e(N) - this has now been fixed (new version
available from the SSC thanks to Kit Baum).

However, I don't know why the output
"Number of sys_id" is produced.

Sophia

Matt Dobra wrote:

```
Dear Statalisters,

I'm having a strange problem when I try to outreg my random effects tobit estimations using GLLAMM. I'm estimating 29 models, and attempting to outreg the results into nine tables. Despite using a very similar outreg command for each one, I am getting something strange. For most of the outregged numbers, the number of observations is incorrect--what it actually reports is the number of random effects. For two of the 29 models, however, it gives the correct number of observations, and adds an extra line to the outreg file (one that I did not call for), called Number of sys_id (which is my i variable), which contains the number of random effects.
Here is my do-file code for one of the strange tables...In the outregged table, the estimates from the fourth dependent varible, zcash, puts the correct number of observations in the observations row and adds a row for "Number of sys_id". The others have the number of random effects in the Observations column. In other words, it looks like this (sorry if the columns don't line up):

1 2 3 4 5 Observations 259 259 259 624 259
Number of sys_id 259

Can anyone tell me what is going on?

xttobit zteq `vars', ll(0) ul(100)
matrix b=e(b)
matrix a=b
local n=colsof(a)
matrix a[1,`n']=b[1,`n'-1]
matrix a[1,`n'-1]=ln(b[1,`n'])
gllamm censorzteq `vars', offset(offzteq) i(sys_id) fam(gauss binom) link(ident sprobit) lv(varzteq) fv(varzteq) from(a) copy adapt dots
outreg using qtbigall.out, nol 3aster se bracket replace

xttobit ztbond `vars', ll(0) ul(100)
matrix b=e(b)
matrix a=b
local n=colsof(a)
matrix a[1,`n']=b[1,`n'-1]
matrix a[1,`n'-1]=ln(b[1,`n'])
gllamm censorztbond `vars', offset(offztbond) i(sys_id) fam(gauss binom) link(ident sprobit) lv(varztbond) fv(varztbond) from(a) copy adapt dots
outreg using qtbigall.out, nol 3aster se bracket append

xttobit zalt `vars', ll(0) ul(100)
matrix b=e(b)
matrix a=b
local n=colsof(a)
matrix a[1,`n']=b[1,`n'-1]
matrix a[1,`n'-1]=ln(b[1,`n'])
gllamm censorzalt `vars', offset(offzalt) i(sys_id) fam(gauss binom) link(ident sprobit) lv(varzalt) fv(varzalt) from(a) copy adapt dots
outreg using qtbigall.out, nol 3aster se bracket append

xttobit zcash `vars', ll(0) ul(100)
matrix b=e(b)
matrix a=b
local n=colsof(a)
matrix a[1,`n']=b[1,`n'-1]
matrix a[1,`n'-1]=ln(b[1,`n'])
gllamm censorzcash `vars', offset(offzcash) i(sys_id) fam(gauss binom) link(ident sprobit) lv(varzcash) fv(varzcash) from(a) copy adapt dots
outreg using qtbigall.out, nol 3aster se bracket append

xttobit zreq `vars', ll(0) ul(100)
matrix b=e(b)
matrix a=b
local n=colsof(a)
matrix a[1,`n']=b[1,`n'-1]
matrix a[1,`n'-1]=ln(b[1,`n'])
gllamm censorzreq `vars', offset(offzreq) i(sys_id) fam(gauss binom) link(ident sprobit) lv(varzreq) fv(varzreq) from(a) copy adapt dots
outreg using qtbigall.out, nol 3aster se bracket append

*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
```*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```